Data Science for Economists

Zhentao Shi

Big data

  • Information technology.

  • Digitalized information.

    • Easy to collect, transport, store, and distribute.
    • Engineering questions.

Data Science in Economics

  • Data science is about the life cycle of data analysis.

  • Economics gives data science purposes

    • Start with a question
    • Locate, clean, and analyze data
    • Interpret, and present data

Why data

  • Scientific research. e.g. Physics.
    • Theory and experimental evidence.
  • Economics research
    • Modeling causality
    • Experimental causality
  • Business analytics
    • Descriptive statistics
    • Prediction
    • Inference

Three questions

  • How much do Msc students earn? (unsurprised learning / descriptive)

  • Is a student’s academic performance related to future salary? (supervised learning)

  • Does higher GPA boost my salary? (Causal question)

Structured data

  • Cross-sectional data
  • Time series
  • Panel data
    • tall and narrow (FRED-MD)
    • short and wide (Health survey)
    • tall and wide (High frequency financial data in market)
  • Tensor data

Unstructured data

  • Texts

  • Graphs

  • Audios

  • Videos

Programming

  • Essential skill
  • Low-level languages: C, Fortran, …
  • High-level languages: Python, R, Matlab, …
  • This course used Python as the primary language

  • R serves as a complement

  • Students can choose their own languages

Coverage

  • Tools for productivity

    • Basic scientific programming
    • Advanced scientific programming
  • General Techniques

    • Simulations, Resampling, Optimization
  • Methods

    • Time series
    • Machine learning, Neural networks

AI

  • Chats

  • Github Copilot

  • Agents

Environment